On Sample-Efficient Offline Reinforcement Learning: Data Diversity, Posterior Sampling and Beyond.
Thanh Nguyen-TangRaman AroraPublished in: NeurIPS (2023)
Keyphrases
- reinforcement learning
- data collection
- synthetic data
- training data
- database
- data sources
- data analysis
- image data
- data sets
- data processing
- high quality
- data mining techniques
- raw data
- data distribution
- computer systems
- probability distribution
- end users
- original data
- data structure
- data quality
- machine learning
- random sampling
- reinforcement learning algorithms
- statistical analysis
- input data
- data points
- data model
- prior knowledge
- learning algorithm
- neural network
- real time